Those are two key conclusions of a study published in Nature this week. Researchers identified eight areas on the human genome that showed a variance in whether people smoked pot, although these areas accounted for 11% of the variance between smokers and abstainers. Many of the genes singled out have not been identified in previous studies.

The study drew on data from 23andMe, the U.K. biobank and the International Cannabis Consortium. Together, they offered researchers a sample size of nearly 185,000 volunteers, five times larger than the largest study to date.

Numerous studies in the past have found a link between cannabis use and schizophrenia, although there has been disagreement on which one, if any, caused the other. This latest study showed that schizophrenia may lead to smoking pot, not the other way around.

“This suggests that individuals with schizophrenia have a higher risk to start using cannabis,” the paper said, referring to its analysis of genetic data. “Our findings may indicate that individuals at risk for developing schizophrenia experience prodromal [that is, early] symptoms or negative affect that make them more likely to start using cannabis to cope or self-medicate.”

The study also found a genetic overlap with cannabis use and other traits such as smoking and alcohol use, schizophrenia, ADHD, and risk-taking, although it did not investigate whether these traits also lead to higher cannabis use.

The study’s authors cautioned that, despite the largest sample size to date, there are other, non-genetic factors that determine cannabis use, such as whether it’s legally available and whether cannabis use is prevalent in a given area. The study also didn’t distinguish between occasional and regular use of cannabis.